1 Uses, cultural significance, and management of peatlands in the Peruvian Amazon: 2 implications for conservation 3 Christopher Schulz a, Manuel Martín Brañas b, Cecilia Núñez Pérez b, Margarita Del Águila 4 Villacorta b, Nina Laurie c, Ian T. Lawson c, Katherine H. Roucoux c 5 a Department of Geography, University of Cambridge, Downing Place, Cambridge CB2 3EN, 6 United Kingdom 7 b Amazonian Cultural Diversity and Economy Research Programme, Peruvian Amazon Research 8 Institute (IIAP), Av. José A. Quiñones km 2.5, Iquitos, 9 c School of Geography and Sustainable Development, University of St Andrews, Irvine Building, 10 North Street, St Andrews KY16 9AL, United Kingdom 11 12 Corresponding author: 13 Christopher Schulz ([email protected]) 14 Department of Geography, University of Cambridge, Downing Place, Cambridge CB2 3EN, 15 United Kingdom 16 17 E-mail addresses co-authors: 18 Manuel Martín Brañas: [email protected] 19 Cecilia Núñez Pérez: [email protected] 20 Margarita Del Águila Villacorta: [email protected] 21 Nina Laurie: [email protected] 22 Ian T. Lawson: [email protected] 23 Katherine H. Roucoux: [email protected] 24 25 Acknowledgements 26 The authors would like to thank the communities of Nueva York and Nueva Unión, Loreto, Peru, 27 for agreeing to participate in this research. Further thanks are due to Sam Staddon and Mary 28 Menton for advice on community benefits, Michael Gilmore on participatory mapping, Harry 29 Walker on doing research in communities, Greta Dargie on peat and peatland 30 characteristics, Eurídice Honorio Coronado, Tim Baker, Jhon del Aguila Pasquel, and Ricardo 31 Zárate on the ecology of the area, and Althea Davies for comments on an earlier version of this 32 manuscript. Funding from the Scottish Funding Council (Global Challenges Research Fund 2017- 33 2018) and the Natural Environment Research Council (ref. NE/R000751/1) is gratefully 34 acknowledged. 35 36

1

37 Uses, cultural significance, and management of peatlands in the Peruvian Amazon: 38 implications for conservation 39 40 Abstract 41 Tropical peatlands play an important role in the global carbon cycle by acting as significant carbon 42 stores. ’s largest peatland complex is located in the Loreto Region of the Peruvian 43 Amazon. Here we present the first study of human relations with these peatlands, including their 44 uses, cultural significance and current management, as well as implications for conservation, based 45 on qualitative research with people living in two riverine rural communities. Our results indicate 46 that peatlands are culturally ambiguous spaces, used mainly for hunting, palm fruit harvesting, and 47 timber, but feared due to the dangers of getting lost, sinking into the ‘sucking’ ground, and being 48 attacked by anacondas and/or mythical creatures. While the difficult terrain and remoteness of 49 peatlands have thus far acted as natural barriers to their destruction through conversion to different 50 land uses, overuse of natural resources is nevertheless a significant concern for people living in the 51 peat-dominated landscape of the Peruvian Amazon, mixed with frustration about the lack of 52 outside support to foster environmental conservation and economic opportunities. We explore 53 how evaluations of the present situation differ across one indigenous and one mestizo community. 54 We identify a range of nascent peatland conservation strategies, including seedling planting to 55 regrow valuable (palm) trees, and the climbing of palm trees for harvesting fruit as opposed to 56 felling them. We argue that peatland conservation could be combined with the development of 57 sustainable management strategies, but that this would require sustained engagement by outside 58 organisations with rapidly growing local communities in these areas. 59 60 Keywords 61 Amazon; conservation; environmental management; peatlands; Peru; Urarina 62 63 1 Introduction 64 Tropical peatlands play an important and, until recently, underappreciated role for the global 65 climate system, due to their capacity to process and store large amounts of carbon (Rieley & Page 66 2016). The largest known areas of peatland in the tropical latitudes are found in Amazonia, 67 particularly in the Loreto Region of Peru (Draper et al. 2014; Lähteenoja et al. 2012), in Southeast 68 Asia (Page et al. 2011), and the Congo Basin (Dargie et al. 2017). While Southeast Asian peatlands 69 have been heavily affected by human uses and degradation (Dohong et al. 2017), South American 70 and African peatlands are still comparatively intact, possibly due to the lower population density, 71 continued availability of more suitable land for agriculture, and the comparatively higher cost of 72 converting remote peatlands for agriculture, among others (Lilleskov et al. 2018; Roucoux et al. 73 2017). 74 Nevertheless, very little is known about local people’s relations with peatlands in the Peruvian 75 Amazon as most research on human uses, management and conservation of tropical peatlands has 76 focused on Southeast Asia (e.g. Medrilzam et al. 2017; Nath et al. 2017; Tachibana 2016; Thorburn 77 & Kull 2015). Globally, people’s relationships with peatlands are shaped not only by material uses 78 of their natural resources (Page & Baird 2016), but also by their cultural status (Byg et al. 2017; 79 Lehtinen 2000; Wilson 2018), which may have implications for their conservation and management. 2

80 Similarly, socio-economic factors play an important role in shaping human-peatland relationships 81 (Dohong et al. 2017; Medrilzam et al. 2017; Tachibana 2016). 82 People living in remote and rural communities are often the principal actors shaping ecosystem 83 management in their surroundings (Álvarez Alonso 2012; Berkes 2004; Fabricius et al. 2007; 84 Waylen et al. 2013), yet their voices and perspectives are seldom heard in wider debates. Here we 85 address this gap by engaging with the views of people living in peatland areas of Loreto, Peru, 86 based on findings from semi-structured interviews and participatory mapping with local 87 community members in two Amazonian communities, and site visits with local guides to 88 neighbouring peatland areas. We examine the material and intangible values local people place on 89 peatlands, the uses they make of them, the forms of management and implications for peatland 90 conservation. In this paper, we define ‘local people’ as ‘people living in rural communities in the 91 immediate vicinity of peatland areas’. Our findings are relevant to decision-makers from local to 92 international scales, seeking to develop appropriate conservation and management strategies for 93 tropical peatlands in the Peruvian Amazon and beyond. 94 95 2 Materials and methods 96 This is the first study to engage with local people’s views on the uses, cultural significance, 97 management, and conservation of peatlands in . We thus followed an 98 exploratory research approach, using multiple qualitative methods. Twenty semi-structured 99 interviews with 27 inhabitants of a small indigenous community of about 150 inhabitants (near the 100 Chambira River, Loreto Region) and 31 interviews with 35 interviewees in a mestizo community1 of 101 about 1,200 inhabitants (near the Tigre River, Loreto Region) were carried out by a team of four 102 Peruvian and British researchers [anonymised for peer review] between March and April 2018 (see 103 Figure 1 for the study location within the wider Pastaza-Marañón Basin). Given the absence of 104 previous research on the topic, we chose to do research in two very different communities. This 105 allowed us to explore the potential role of cultural differences, as well as variations in factors such 106 as community size, history, integration into wider Peruvian and economy, among others, 107 for people’s relations with surrounding peatland areas. Both communities had been visited by 108 members of the research team prior to our fieldwork, which we believe helped to establish the 109 necessary trust to conduct effective research there. 110 [Insert Figure 1 around here] 111 The indigenous community is exclusively populated by members of the Urarina indigenous nation 112 whose ancestors are likely to have lived in the area for centuries, although the exact location of the 113 current community was only chosen about 30 years ago. Available infrastructure consists of a 114 primary school and a church (each with their own small generator for electricity), a public speaker 115 system, and a radio system to communicate with neighbouring communities. In contrast, the mestizo 116 community was founded 85 years ago by (former) rubber tappers from a distant community 117 elsewhere in the Peruvian Amazon, and is now inhabited by descendants of immigrants from many 118 different parts of the Peruvian Amazon. Culturally, its inhabitants could thus also be classified as 119 ribereños (Chibnik 1991), i.e. mestizos with cultural roots and significant environmental knowledge of 120 the Amazon who have lived in the area for generations. Available infrastructure consists of 121 generator-powered electricity for most of the community, a primary and secondary school, two

1 Mestizo - Spanish for mixed race, an ethnic category that can be traced back to the times of Spanish conquest, which nowadays has mainly cultural connotations about differences in ancestral knowledge and practices. 3

122 churches, a community hall, a public speaker system, mobile phone coverage, a health post, a police 123 station, two concrete pavements, two (modest) hotels, and a number of corner shops. Neither 124 community had running water or sanitation infrastructure. 125 Our sample of interviewees ranged in age from 18 to 80 years old, with 34 male and 17 female 126 respondents who were interviewed in Spanish (with the exception of seven female interviewees in 127 the Urarina indigenous community who were interviewed in Urarina with the help of a local 128 translator). Most male respondents were small-scale subsistence farmers who also engaged in 129 hunting and fishing. A majority was also active in the seasonal trade of aguaje palm fruit (Mauritia 130 flexuosa), chonta or huasaí palm hearts (from Euterpe precatoria), and a few agricultural products such 131 as manioc (processed as fariña). Many had worked for oil companies and sold timber in the past 132 (timber harvesting was ongoing in the mestizo community). A minority had additional professions 133 with an associated monetary income, e.g. as school teacher, carpenter, or small-scale retailer. 134 Female interviewees also worked in small-scale subsistence agriculture and fishing, and were 135 primarily in charge of childcare, cooking, collecting firewood, washing clothes, and production of 136 textiles. In the mestizo community, some female interviewees also worked in the trade of aguaje palm 137 fruit and in retail. 138 Semi-structured interviews covered four main themes, namely (1) natural resource use (e.g. game 139 species, palm fruits, timber trees or fish); (2) classification of the environment surrounding the 140 communities into different ecosystems (see Authors 2019 for an overview of Urarina indigenous 141 ecosystems, and Halme & Bodmer 2007 for an overview of some of the ecosystems recognised by 142 mestizos in the northern Peruvian Amazon); (3) cultural and mythological importance of certain 143 ecosystems, especially those likely to have peat-rich soils; and (4) environmental governance, 144 including past, present, and potential future strategies for the conservation and management of 145 areas surrounding communities, again with a particular focus on peatland ecosystems. All 146 interviews were recorded and transcribed in Spanish, and analytical categories within the interview 147 transcripts were coded with NVivo 11 to facilitate the qualitative analysis. 148 As part of the semi-structured interviews, research participants were asked to locate resources, 149 locally defined ecosystems, and areas that they personally travel to on A3-size maps of the areas 150 surrounding the two communities. Additionally, one community-scale workshop was conducted 151 per community, with about 50 attendants each, to perform a participatory mapping exercise with 152 large A0-size maps (which showed, as a starting point, only the location of the community and the 153 main rivers and streams). The benefits of participatory mapping for safeguarding traditional 154 ecological knowledge, strengthening territorial rights of local communities, and supporting 155 environmental conservation have been well documented (Gilmore & Young 2012; Ramirez- 156 Gomez et al. 2013). Moreover, this method can serve as a vehicle to start conversations about uses 157 and management of the environment, beyond merely locating resources and areas on a map. Not 158 least, participatory mapping is also an enjoyable activity for research participants, which can serve 159 as a tool for education within communities when knowledgeable community members share their 160 experiences about the local geography (Young and Gilmore 2013). Further information on peatland 161 uses and management was gathered during a total of six site visits to neighbouring ecosystems and 162 areas likely to be peatlands with local community members in both locations. 163 164 3 Uses, cultural significance, and management of peatlands in the Peruvian Amazon 165 Given that respondents were not familiar with peat as a type of soil, we relied on various proxy 166 indicators to identify peatland areas around communities, such as the colour of soils and their 4

167 porewater (i.e. dark brown to black); ‘sinkiness’ of the ground (i.e. area where one sinks in easily); 168 amount of water in the soil (i.e. permanently waterlogged areas); and, occasionally, vegetation 169 appeared to be an indicator (i.e. areas where trees are shorter than usual; where there are only 170 grasses and sedges). We were also able to visit local ecosystem types whose descriptions matched 171 those characteristic of peatlands, based on information gained from the individual semi-structured 172 interviews and participatory mapping exercises, and presence/absence of peat was further verified 173 by probing with a pole. 174 In the indigenous community, two local ecosystem types known as jiiri and alaka are likely to be 175 typically associated with peat (see Authors 2019). In the mestizo community, the waterlogged areas 176 likely to be peatlands were often identified as ‘ugly’ areas, classified into aguajal chupadera (Mauritia 177 flexuosa palm swamp with soft and wet ground), aguajal raizal or aguajal champal (Mauritia flexuosa 178 palm swamp with comparatively firm ground where roots cover the soft and waterlogged soil), 179 aguajal varillal (Mauritia flexuosa palm swamp with many short and thin trees; note that varillal by 180 itself also exists as a vegetation category in common usage, and can be translated as ‘pole forest’ 181 sensu Draper et al. 2014 or Laumonier 1997), and piripiral (sedgeland with very soft ground and 182 lacking trees). Use of terms varied between individual respondents, with aguajal chupadera (i.e. 183 ‘sucking’ aguajal) being the most common term to refer to areas that we identified later as having a 184 peat substrate. Aguajal chupadera and piripiral were regularly identified as ‘particularly ugly’ areas (i.e. 185 “allá es feísimo”, in Spanish), given the great difficulty walking on the soft, waterlogged soil typical 186 of peatlands. In contrast, some respondents described areas with firm ground as ‘beautiful areas’ 187 (i.e. “allá es lindo”, in Spanish). Piripirales (sedgelands) were also associated with ‘dead lakes’ (cochas 188 muertas) in the mestizo community, with a special cultural status (see section 3.2 below). For an 189 overview of local peatland terminology, uses, cultural significance, and current management and 190 conservation strategies in the two studied communities, see Table 1. 191 [Insert Table 1 around here] 192 193 3.1 Uses of peatlands 194 People in the Peruvian Amazon do not use peat itself. Nevertheless, peatland areas provide 195 livelihoods and economic income to local people indirectly, through the plant and animal resources 196 that can be found there. A number of resources are collected primarily for subsistence and personal 197 consumption, most notably meat from terrestrial mammals (e.g. tapir, peccary, agouti), monkeys 198 (e.g. howler, squirrel monkey, monk saki), reptiles (e.g. caiman, tortoise), and birds (e.g. Spix’s guan, 199 great tinamou). While these animals can be found in peatland areas, most can equally be found in 200 non-peatland areas, with tapirs and caimans most commonly mentioned as typical game species of 201 peatland areas specifically. There was some limited intra-community trade of meat in the mestizo 202 community, but not in the (much smaller) indigenous community. Further non-commercial 203 resources found in peatlands, e.g. the leaves of the shebón palm tree (Attalea butyracea) used for 204 traditional roofs (although these are increasingly replaced by corrugated metal roofs), and the fibre 205 of aguaje palm trees (Mauritia flexuosa) used by members of the indigenous community to produce 206 traditional textiles, were also important. Some respondents also mentioned limited harvesting of 207 palm fruit from several palm tree species for personal consumption, such as ungurahui (Oenocarpus 208 batahua) or aguajillo (Mauritiella armata) (see also Smith 2015). As with animals, these plant resources 209 are not exclusive to peatland areas, although relatively common there. 210 Economically important products include fruit from the aguaje palm tree (Horn et al. 2018), and 211 palm hearts from the huasaí or chonta palm tree (Euterpe precatoria) (Paniagua-Zambrana et al. 2017), 5

212 which members of both communities regularly harvested to sell to travelling traders. In the mestizo 213 community, there was a simple commercial value chain whereby a network of intermediaries would 214 buy these products from other community members and sell them in bulk to traders from Nauta 215 and Iquitos (the two closest towns in the northern Peruvian Amazon). Due to the strong economic 216 importance of this trade, one community member called their community by the nickname “The 217 capital of aguaje”, although aguaje trade occurs on similar scales elsewhere in the region (Horn et al. 218 2018). Both aguaje fruit and chonta palm hearts are typically harvested by felling palm trees during 219 the harvesting season, which lasts several months of the year. While aguaje palm trees do not need 220 peat to grow, they are extremely common in areas with a persistently high water table and frequently 221 form monodominant stands (locally known as aguajales, see Endress et al. 2013), which often overlie 222 peat in this area (Freitas Alvarado et al. 2006). 223 Finally, wood and timber products were also harvested from peatland areas, mostly for personal 224 use, and in the mestizo community, for trade and monetary income as well (in the indigenous 225 community, timber harvest and trade ceased about five years previously). The most frequently 226 mentioned species was cumala (Virola sp.); further timber species mentioned were e.g. moena (which 227 may refer to several species of the Lauraceae or laurel family), shiringa (Hevea brasiliensis; rubber tree), 228 lagarto caspi (Calophyllum brasiliense), although several interviewees commented that trees from 229 peatland areas might have less durable wood of lower quality than from non-peatland areas. Beyond 230 these timber tree species, stems of smaller trees typical of peatland palm swamps and pole forests 231 were occasionally used as poles for construction (especially in the indigenous community), e.g. 232 punga (Pachira brevipes). Remo caspi (Aspidosperma rigidum) was mentioned as a source of firewood in 233 the mestizo community, despite traditionally being favoured for making oars (remo = oar). It should 234 be noted, however, that remo caspi was more strongly associated with low-lying areas in general 235 (bajiales), which may not necessarily be peatlands. 236 237 3.2 Cultural significance of peatlands 238 While peatlands themselves are not recognised as such by members of the two communities, the 239 ecosystems typically associated with peat had a special cultural status among indigenous and mestizo 240 respondents alike. In the Urarina indigenous community, the jiiri and alaka [peatland] ecosystems 241 were considered to be the home of a mythical creature, the guardian spirit Baainu, who may trick 242 people into losing their way (see Authors 2019 for a detailed discussion), with some similarities to 243 forest guardian spirits elsewhere in the Amazon more broadly (see e.g. Smith 2015). They were also 244 of special cultural importance as the source of aguaje fibre for textile production. Aguaje textiles play 245 a central role in Urarina cosmovision and their creation myth, which includes an element in which 246 a ‘wise’ woman is identified by her ability to weave aguaje palm-fibre cloth (Dean 1994). In the 247 mestizo community, many guardian spirits were known, too, most commonly under the name 248 Yashingo (other names used were Chullachaqui, Sacharuna, Shapshico, Yacumama, or simply madre 249 [mother] or dueño [owner]), but most of these were not specific to peatlands as a broader ecosystem 250 category. Instead, some madres or dueños inhabit specific individual lakes and their surroundings, 251 which is a common pattern in Amazonia more broadly (see e.g. Mezzenzana 2018; Ricopa Yaicate 252 2009). 253 Some of these lakes belonged to a category locally known in the mestizo community as cocha muerta 254 (dead lake). Based on respondents’ descriptions, it appears that the areas around ‘dead lakes’ could 255 have peat substrates. These were described as sedgelands (piripirales) or grasslands (hierbales) with 256 “very ugly” soil, where one can sink in to an extent that it poses a risk of drowning. Trees were 257 either absent or much shorter than usual, with the most common tree species being renaco (Ficus sp. 6

258 or Coussapoa sp.). In the wider surroundings, aguaje might be found as well. All respondents familiar 259 with dead lakes also noted an abundance of aquatic plants such as rayabalsa (Montrichardia arborescens) 260 and huama (Pistia stratiotes), i.e. their surface is usually covered by vegetation. The water of dead 261 lakes was described as follows: “The water is brown […], like [Coca Cola], extremely ugly, like 262 when you mix a soft drink with milk, […] a very thick water” (male respondent, mestizo community, 263 27 years). Contrary to their name, dead lakes are actually full of fish, and game species such as river 264 turtles, tapirs, and monkeys can comparatively easily be found nearby, which is why a minority of 265 settlers still visit them, despite being fearful. 266 Theories about the origin of the name ‘dead lake’ varied, with some attributing it to the lack of 267 trees, and others citing their calmness which resembled a genuinely dead lake. It seems possible 268 that the name for these comparatively remote lakes and ponds is as much related to their cultural 269 status as taboo areas inhabited by particularly powerful and irritable ‘mothers’ and ‘owners’, as to 270 the actual risk of death by attacking anacondas or caimans, or sinking into the soft ground. The 271 ‘mothers’ of dead lakes can be angered easily by any sort of noise, such as from rifles used for 272 hunting or boat engines, and will retaliate with instant thunderstorms. Such ‘mothers’ often take 273 the shape of anacondas, and respondents noted being particularly fearful of anaconda attacks in 274 dead lakes. Reportedly, these are also the home of giant black caimans with lengths of up to 12 275 metres. Also, the renaco trees growing near dead lakes possessed a special mythological role. Several 276 respondents reported that felling a renaco would lead to the death of the person in question and 277 their entire family, which effectively protected renacos from the timber trade. The degree to which 278 community members believed in supernatural phenomena differed widely, and descriptions of dead 279 lakes were not uniform either. Interestingly, such beliefs were sometimes held in parallel to 280 Christian beliefs, introduced through activities and maintained by two evangelical 281 churches in the mestizo community. 282 In the mestizo community, other types of ecosystems likely to be peatlands (i.e. the various types of 283 aguajal mentioned above) did not appear to have a similar cultural significance for local mythology 284 as dead lakes. Nevertheless, through their strong socio-economic importance as the source of aguaje 285 fruit, they still had a great influence on everyday life in the community, and in this way, on local 286 . Indeed, the aguaje palm tree could be understood as a cultural keystone species in the sense 287 proposed by Garibaldi and Turner (2004), with great importance for ecology and local culture alike. 288 For example, the aguaje harvest is a very important element of the community’s seasonal calendar 289 and, at harvesting times, would dominate community life. Nevertheless, most respondents also 290 emphasised (and maybe lamented) the physically difficult environment in which aguaje grows. This 291 difficult environment tends to shape peatlands’ perceptions as culturally ambivalent spaces in Peru 292 (Authors 2019) as elsewhere around the globe (Boaden 1981; Byg et al. 2017; Lehtinen 2000; 293 Wilson 2018). 294 295 3.3 Current management and conservation of peatlands and potential threats 296 Neither of the two communities studied had specific management practices in place for peatlands. 297 If anything, the cultural taboos surrounding dead lakes in the mestizo community, and fear of the 298 Baainu in the Urarina community, may act as (weak) indirect incentives for environmental 299 conservation of some peatland areas specifically. Nevertheless, both communities had local 300 agreements in place with a view to conserving their natural resources, including those in peatlands, 301 which were complied with to varying degrees.

7

302 The most effectively implemented strategy to conserve natural resources concerned their 303 protection against outsiders. Any non-members of the community caught fishing, hunting, or 304 harvesting timber inside the community territory would be sanctioned, either with verbal warnings, 305 fines, temporary detention, or confiscation of their equipment and boat engines. In the larger 306 mestizo community, a police post was in place to take care of any potentially illegal activities by 307 outsiders, mostly by monitoring river traffic day and night. The indigenous community also had 308 agreed on a ban on using poison for fishing, which is a very effective, but unsustainable traditional 309 fishing method. While this is illegal according to Peruvian law, people in both communities were 310 generally only aware of the regulations agreed within their own local governance systems. 311 These community agreements were framed by respondents as the result of community deliberation 312 and self-governance, but also appeared to be strongly related to the initiative of incumbent local 313 leaders, who in turn may have been influenced by interacting with state authorities, NGOs, and 314 other institutions (see e.g. Cossío et al. 2014). Both communities used a dual governance system 315 typical for indigenous communities, in which indigenous leaders (officially known as apu and vice- 316 apu) and state-recognised leaders (known as teniente gobernador and agente municipal) would be elected 317 by community members and govern collaboratively. In the indigenous community, there was also 318 a madre indígena (‘indigenous mother’) who represented women’s concerns specifically. This position 319 has been developed relatively recently as a result of state and other outside demands for better 320 representation of women’s interests and needs in indigenous communities. These demands were 321 passed to communities via indigenous federations, which in turn unite representatives of several 322 communities within a certain geographical area. However, most major decisions would be taken by 323 the community as a whole, and both communities had a formal register of decisions taken, signed 324 by all attending community members. The local police post of the mestizo community was 325 established due to lobbying of the state authorities by local leaders, following an incident in which 326 a group of criminals murdered a member of the community and injured two others (the reason for 327 the incident remained obscure, but might have been related to drug trafficking; see Perú21 2012). 328 Both communities had also formally agreed to limit their use of natural resources through the local 329 governance system, including fish, meat, and timber, but compliance was mixed at best. In the 330 mestizo community, a monthly limit on selling fish had been agreed, but different respondents cited 331 different figures (between 20 kg and 50 kg/month/family), and many openly admitted that non- 332 compliance with this rule is the norm. This was mostly attributed to the lack of alternative sources 333 of livelihoods, the need to provide for one’s family, and, when referring to other people’s behaviour, 334 their indifference to environmental conservation (in the sense of a symptom of the broader 335 undesirable character trait of carelessness). In the indigenous community, a former local authority 336 had brokered a temporary agreement among community members during his tenure, to stop all 337 commercial timber harvesting activities to let trees regrow (possibly related to the presence of an 338 environmental NGO in the community at the time, see Cossío et al. 2014), which seemed to be 339 complied with. However, some respondents suggested that this was entirely due to the recent 340 appearance of alternative economic opportunities, in the form of employment in the maintenance 341 of a local oil pipeline corridor (conversely, the former local authority cited the agreement as a local 342 environmental management success). Thus, experiences in both communities highlight the need 343 for diversifying conservation strategies beyond simple resource use bans, which may also 344 disproportionately affect the most marginalised community members, given that these typically 345 lack alternative livelihood strategies. 346 Despite the problems with community-level environmental management strategies, almost all 347 respondents noted the need to protect natural resources to ensure the sustainability of local 348 livelihoods. Concerns about resource depletion and ecological change, including in areas which we 8

349 identify as likely to be peatland areas, were common in both communities but much more 350 pronounced in the mestizo community. While in the indigenous community some respondents 351 commented on the disappearance of commercially valuable timber species such as mahogany 352 (Swietenia macrophylla, known as caoba in local Spanish) or cedro (Cedrela odorata) in the past few decades, 353 there was no sense that survival of the community and traditional livelihood strategies (hunting, 354 fishing, small-scale farming) were at stake. In the mestizo community, however, there was a sense of 355 doom about disappearing natural resources, i.e. fish, game species, and aguaje, which was often 356 related with the (possibly idealised) abundance of the past: 357 “Here we used to have a lot of fish… if you pointed a torch at the river at low water 358 levels, the eyes of the caimans were shining like electric light […] just here at the mouth 359 of the Tigrillo River, you could see the fasacos [Hoplias malabaricus, a local fish species] 360 on the river margins, we did not care, they appeared to be [as abundant as] wooden 361 sticks. […] Same thing with the aguaje, you used to find it right here, but nowadays 362 people have to walk the whole day to harvest two, three, four bags… they started 363 destroying right here […] and they continue destroying until today.” – male respondent, 364 mestizo community, 72 years 365 Overall, overuse of natural resources, caused by the lack of alternative livelihood strategies and 366 population growth, was the most common threat identified by members of the mestizo community. 367 While community-level conservation strategies appeared to be difficult to enforce, individual 368 members of the mestizo community had developed their own environmental management strategies 369 to address unsustainable aguaje harvesting practices (this also echoes comments e.g. by Waylen et 370 al. 2013 about the importance of recognising intra-community differences in conservation). This 371 has important implications for peatland conservation, given that aguaje mostly grows in peatlands 372 in the Pastaza-Marañón Basin, and degradation of aguaje stocks may reduce the carbon storage 373 function of aguaje-dominated peat swamp forests (Bhomia et al. 2018). These can be roughly 374 grouped into three broad themes: (1) climbing, rather than felling, aguaje palm trees for harvesting 375 the fruit; (2) planting of aguaje seedlings to restore depleted areas; (3) identification of alternative 376 monetary income strategies. The first two strategies are closely related given that at present, 377 climbing palm trees is only possible indirectly, where other trees are planted next to aguaje palm 378 trees to act as a ladder. Other trees have branches that can be used for climbing, unlike the single- 379 stem aguaje palm trees, which also grow very tall when mature, making them impossible to climb 380 without suitable specialist equipment (not known in the community prior to our fieldwork). 381 Considering time spans of between seven to nine years for aguaje to bear fruit for the first time 382 (Gonzáles Coral & Torres Reyna 2010), planting seedlings requires a relatively long-term vision, 383 but was portrayed as worth the effort by the small minority of community members who engaged 384 in it: 385 “I have planted [aguaje palm trees] right here [in my orchard]. Now this takes away my 386 stress when sometimes I don’t have money, I sell some [aguaje fruit] and I get money. 387 I tell people from here, if you don’t have [money], it’s because you don’t plant [aguaje], 388 it’s really because you don’t want to.” – female respondent, mestizo community, 69 years 389 Similarly, felling of aguaje palm trees was occasionally described as an income strategy of last resort: 390 “I feel sad felling aguajes […], because sometimes when you fell an aguaje […], 391 sometimes it is full of fruit, and the next year it’s not going to be there. […] That’s why 392 I hardly go… only when I don’t have work here I might sometimes go there [to harvest 393 aguaje].” – male respondent, mestizo community, 33 years 9

394 The same respondent also noted having trained as a carpenter to avoid relying on harvesting natural 395 resources for his monetary income, and had planted several aguaje palm trees in his orchard so that 396 his children would have a sustainable supply of aguaje in the future. Other parents were hoping for 397 their children to study and train as teachers, nurses, and other professions so that they would not 398 have to rely on the dwindling natural resources of their community. 399 In the indigenous community, there had been some planting of seedlings of timber species by an 400 environmental NGO in the past (see Cossío et al. 2014: 11-12), which was generally welcomed by 401 community members at the time. They had also received some training in environmental 402 management. Yet, not long after the NGO left, replanting activities were discontinued, indicating 403 a need for sustained support if communities are to obtain a benefit from external involvement in 404 resource management (Davies & White 2012). 405 Despite the numerous obvious benefits of planting (aguaje) seedlings, this strategy also comes with 406 a number of challenges. Planted seedlings would be considered as economically valuable 407 investments by those planting them, potentially requiring the allocation of more formal land use 408 rights if this was taken up on a larger scale. At present, community members are typically free to 409 cultivate any unoccupied land they find available, subject to approval by the community. Large- 410 scale monoculture plantations of aguaje would likely not be viable due to the species’ susceptibility 411 to numerous pests and diseases (Smith 2015). And not least, it is a dioecious palm species, i.e. it 412 takes years to find out whether seedlings are male or female, a problem reported by several 413 interviewees. Expert guidance suggests that the best way to deal with this uncertainty is to adopt a 414 long-term perspective, which involves gradually replacing male trees with new seedlings, which will 415 then increase the share of female trees over time (Gonzáles Coral & Torres Reyna 2010). Further 416 recommendations for aguaje management include felling older trees, which may not be as 417 productive, to accelerate the growth of younger, fruit-carrying trees; as well as letting some fruit- 418 bearing branches fall to the ground, to allow natural dispersal of seeds via aguaje-consuming animals 419 (Aquino 2005). 420 421 4 Implications for conservation 422 4.1 Cultural and ecological degradation as a related process 423 As noted e.g. by Pröpper and Haupts (2014), culture and ecology are often closely linked among 424 people living in remote rural settings, to an extent that culture does not exist separately of the 425 natural landscapes that people rely on for their (subsistence) livelihoods. It follows that ecological 426 degradation may go along with cultural degradation and vice versa. In the two studied communities 427 (i.e. both indigenous and mestizo), such processes seem evident as well. Several respondents 428 commented that formerly resource-rich areas used to be populated by numerous ‘mothers’ and 429 spirits, which might attack humans; see for example, the following vivid description of the turbulent 430 past of a major lake near the mestizo community (which here emphasises spiritual and mythological 431 aspects of culture): 432 “You could not go to that lake. You arrived there and would hear a loud noise, thunder, 433 rain, you could not enter, its water boiled. That is what my grandfather told me […]. 434 ‘Son, when I arrived there, immediately lightning would strike, it was a very rough lake.’ 435 The Supay [River] as well, because of the anacondas. That’s what my grandfather told 436 me when I used to walk around with him. The entire area was very rough.” – male 437 respondent, mestizo community, 36 years

10

438 However, such fighting back by ‘mothers’ and spirits via (super-)natural phenomena (sudden 439 thunderstorms; anaconda attacks) was said to disappear as more and more humans visited an area 440 to hunt and fish. While the noise of boat engines would initially anger the ‘mothers’, persistent 441 disturbance would eventually make them flee, just like game species and fish. A fully degraded area 442 would be empty of spiritual beings as well, suggesting a particularly close link between ecology and 443 culture. 444 In the indigenous community, there was evidence for such linkages as well. For example, some 445 younger people would consider traditional ecological knowledge, as well as knowledge of cultural 446 traditions and customs, as being beyond their expertise, and would try to redirect our queries to 447 the two oldest male community members. Other cultural-ecological traditions were still maintained 448 to some degree, for example the practice of using aguaje fibre for textile production. This is of 449 strong cultural importance to the Urarina, given e.g. the central role that aguaje textiles occupy in 450 everyday culture and their creation myth (Dean 1994), as mentioned in section 3.2. 451 Nevertheless, while under threat, cultural and ecological knowledge was still better conserved in 452 the indigenous community than in the mestizo community, as evidenced by the continued practice 453 of producing aguaje textiles among the Urarina, for example (see further examples below). This is 454 maybe not surprising, as their ancestors had lived in the area for centuries and most community 455 members would not usually leave the area, except for community leaders and their families. Family 456 links were exclusively with other Urarina communities in the region. In contrast, the mestizos had 457 been present for 85 years at most and their ancestors originated from many different areas (as far 458 away as Portugal) and had family all over Peru (and beyond, with one respondent mentioning her 459 cousin visiting from the US). In the 31 interviews conducted there, 53 different towns and 460 settlements were mentioned where people had travelled or knew someone (without this being an 461 explicit focus of our research). 462 Such different levels of cultural and ecological knowledge have significant environmental 463 management implications, including for peatland areas, as also noted by Paniagua Zambrana et al. 464 (2007). Where cultural traditions erode, for example through integration into mainstream Peruvian 465 society, this may create challenges for environmental conservation, whereas the conservation of 466 traditional practices can be beneficial to prevent ecological degradation. For example, it appears 467 that the Urarina tradition of moving community locations from time to time (described also by 468 Kramer 1979) is a suitable strategy to cope with resource ‘depletion’ in a certain area, which is still 469 being practiced today. As mentioned above, the indigenous community we visited had only been 470 in their current location for about 30 years. And earlier in 2018, a neighbouring community had 471 also changed their location, in this case from a smaller tributary to the shores of the much larger 472 Chambira River. In fact, Urarina hunters or fishermen do not appear to be particularly selective 473 when choosing which animal or fish to catch. Thus, one could imagine a similar dynamic of gradual 474 disappearance of natural resources to occur over the longer term, as described in the mestizo 475 community, which would then be mitigated by moving location. 476 While from an ecological point of view, moving an entire community may be beneficial, it can 477 create legal problems. The Peruvian state is slow to recognise such moves and land titles are usually 478 fixed to a certain territory – indeed the Urarina community we visited was still lobbying to have 479 their new location officially recognised through an enlargement of their territory, which still only 480 covers the area around their previous location. The state may also be reluctant to concede 481 comparatively valuable non-wetland territory to communities, given that the Urarina usually settle 482 in small dry patches surrounded by wetlands, including peatlands. Many other Urarina communities

11

483 in the region do not have any legal recognition at all and are thus simply ignored by the state 484 (Walker 2013). 485 Strong solidarity between community members, expressed for example by sharing meat and fish 486 (Álvarez Alonso 2012) is another important traditional Amazonian strategy to cope with food 487 insecurity, which is inherent to a subsistence lifestyle that is based on hunting and fishing. These 488 practices were present in the indigenous community where no shops existed and monetary 489 exchange was only practiced with non-community members, if at all. In the mestizo community, 490 sharing of food and natural resources also existed to help those less well off, but a clear tendency 491 towards market transactions was evident from the existence of several shops, as well as from 492 comments about trade between community members: 493 “Here we have a tariff for selling meat to each other, it is very well established. We 494 have our norms well thought out. For example for the meat, we charge each other 4 495 soles when it is fresh, and 4.5 soles when it is dried.” – male respondent, mestizo 496 community, 57 years 497 Similarly, if the inhabitants of the mestizo community were to disperse and move to different 498 currently uninhabited areas, it is possible that they could start subsistence hunting, farming, aguaje 499 harvesting, and fishing afresh, in order to let populations recover at their current location. Yet, we 500 would argue this is unlikely to happen as they do not share the same traditions as their indigenous 501 counterparts. Going forward most inhabitants of the mestizo community seem to favour integration 502 into the wider Peruvian economy over continuing with subsistence livelihoods. The practical 503 challenges of achieving this type of economic development in the mestizo community could 504 potentially generate and exacerbate intra-community conflict, as well as an increasing sense of 505 desperation among local people and strong expectations of support from visiting outsiders, state 506 institutions, NGOs, business, and research (see following section). 507 508 4.2 Peatland science and implications for conservation 509 As might have been expected, our research has shown that peatlands as a particular landscape 510 category are not of concern to local people in the Peruvian Amazon. Even among university- 511 educated elsewhere in the country, only some specialists would be familiar with the 512 Spanish term for peatlands (turberas), because peat has had no historic uses in Peru, unlike in many 513 other countries (Braadbaart et al. 2012; Cruickshank et al. 1995; Gapsalamov 2015). Nevertheless, 514 our research has also shown that peatlands are sufficiently distinct and recognisable to potentially 515 allow local or collaborative management of these areas, and that there are local terms for 516 ecosystems that are typically or frequently associated with peat. 517 Respondents in both communities suggested that these areas could be protected by temporary bans 518 on the use of certain natural resources, the planting of seedlings of commercially valuable (palm) 519 trees, including aguaje and cumala, and the creation of alternative economic opportunities. While the 520 origin of the ideas for these conservation strategies is unclear, they may have been informed to 521 varying degrees by previous interactions with environmental NGOs, state development institutions, 522 commercial actors, and academic researchers, who have been active in the region for decades 523 (Schleicher et al. 2017; Zinngrebe 2016), not least in the very large and relatively near Pacaya- 524 Samiria National Reserve (see Figure 1; Kilbane Gockel & Gray 2009). In the mestizo community, 525 several respondents also suggested that commercial agriculture could replace subsistence farming, 526 fishing, and hunting in the future once local fish and game species were depleted. Given the poor

12

527 soils and seasonal flooding regime, this may not prove to be practical, however, and experience 528 elsewhere in the Amazon also shows that such shifts from subsistence to commercial agriculture 529 often lead to displacement and marginalisation of the original small-scale farmers (Ioris 2017). 530 The desire for alternative monetary income sources was especially pronounced in the mestizo 531 community where, overall, there was clearly a joined-up understanding that environmental 532 conservation was inseparable from the underlying issue of insufficient economic opportunities for 533 community members (even if levels of concern for environmental conservation differed 534 considerably between respondents). This was to be achieved through combining conventional 535 economic development with better assistance from state authorities, NGOs, and international 536 donors. The most frequently mentioned option to create jobs was to build a processing factory for 537 aguaje fruit in the community, an idea which had reportedly been mooted by Korean investors in 538 the past, and remained popular even after the investors disappeared. It would be essential, however, 539 to combine this with verified sustainable harvesting practices to avoid increasing pressure on 540 already degraded aguajales. Nevertheless, processing of sustainably sourced aguaje fruit would relate 541 well with the idea of peatland conservation, given that aguaje is common in peatland areas in the 542 Peruvian Amazon (Freitas Alvarado et al. 2006). 543 Conservation of peatlands is a science-driven endeavour, in the sense that without the recent efforts 544 to quantify carbon storage capacities of Peruvian tropical peatlands (Draper et al. 2014; Lähteenoja 545 et al. 2012), peatland conservation would not be on any environmental management agenda in Peru. 546 This makes peatland scientists, voluntarily or not, one of the key actors in any debates on peatland 547 conservation, both with local, national, and international stakeholders, as well as with local people 548 in peatland areas. The need for peatland scientists to reflect on peatland conservation has also 549 motivated the present study. Having identified the magnitude of carbon storage in Peruvian 550 peatlands (Draper et al. 2014; Lähteenoja et al. 2012), lobbying for their conservation to avoid 551 negative implications for the global climate seems to be a logical next step, even if it is often difficult 552 to be heard by relevant decision-makers. 553 Among local community members, Peruvian and especially international scientists are often 554 perceived as potential development workers, a role that they are then forced to engage with, if only 555 to explicitly reject it. For example, one respondent (a former community leader) suggested the 556 introduction of what, for scientists, would be termed as a type of Payments for Ecosystem Services 557 (PES) scheme (which already exist elsewhere in the Pastaza-Marañón Basin, see e.g. Roucoux et al. 558 2017 for a summary of a carbon-based conservation initiative of the Green Climate Fund2). In a 559 long narrative indicating his experience with the wider framing of development and conservation 560 initiatives in the region, he explained the need for developed countries such as Scotland to pay 561 community members for conserving peatlands and palm swamps: 562 “Previously, there was a project here in Peru, […]. We heard that foreign countries 563 were sending money to every family of a village, and they all had a limited area, of 564 about ten or twenty hectares to look after, let’s say, where they had to plant trees. But 565 there were conditions for having this salary, planting and conserving trees, […], and in 566 return, they went to the bank to get their payment […]. We heard about this [project] 567 in Nauta, and we asked ourselves, why did they not come to Nueva York [i.e. the 568 mestizo community]? […] It is always like that, you go there [to the regional authorities 569 in Nauta] and they say a project is coming, but they cheat you and it actually doesn’t 570 come. […] And then [when you return to Nauta] the project is completed already. […]

2 More information is available here: https://www.greenclimate.fund/projects/fp001 13

571 Then they say, we are going to extend the project, but pucha, we don’t know what’s 572 happening, we haven’t seen anybody. […] Several communities were going to be part 573 of this, but nothing happened. 574 […] 575 Scotland is a developed country, right? […] I would like to suggest, […] this idea that 576 I have, maybe you could take it to your country: a project [like the above], […] about 577 reforestation, […] that way we would have had a way to sustain ourselves, we wouldn’t 578 be cutting neither the aguaje palm trees nor the chonta palm trees, nor the timber trees, 579 all that. It might already not be for our own benefit, but for the future generations that 580 are coming.” – male respondent, mestizo community, 53 years 581 Such (perhaps overly ambitious) demands for foreign intervention are not uncommon when 582 foreign researchers visit remote rural communities (see e.g. Staddon 2014; Townsend 1995). 583 Nevertheless, these are often difficult to navigate on the ground even when prior consultation and 584 the parameters for engagements and desired outcomes for communities have been agreed between 585 all parties. An example of this occurred in a previous scoping visit for this project when the 586 researchers suggested that the project could facilitate a palm tree climbing workshop with 587 community members, in which they would learn how to climb aguaje palm trees, rather than cutting 588 them, for harvesting fruit. (This approach has numerous ecological and economic benefits, and 589 strongly enhances sustainability of aguaje fruit trade [Smith 2015]). At the time of prior consultation, 590 in a meeting attended by 50 adults (21 women/29 men), the community had seemed strongly 591 supportive of this idea. When it came to implementing the workshop, however, attendance was 592 very low (it proved popular mostly with children and adolescents). Of course, low turn-out in such 593 contexts usually has a range of explanations (Cheng & Mattor 2006; Davies & White 2012; 594 Messerschmidt 2007). In this particular case, timing may well have been an issue as people were 595 rushing to save their manioc harvests from being flooded by unexpected rises in the river at that 596 time. 597 Yet, it is also important to recognise the legacies of the experience of past projects on undermining 598 trust. Local people reported their negative memories of NGOs, government entities, businesses, 599 and others disappearing without a trace after announcing grand development plans (e.g. building 600 an aguaje processing factory). They mentioned their time being wasted by ineffective programmes 601 like teaching vegetable production and then handing out low quality seeds unsuitable for local soils. 602 Others cited how their money had been stolen in different ways – for example a monetary 603 investment was required to participate in an initiative but the programmes were never completed 604 or their payments disappeared via corrupt channels including, in some cases, within the community. 605 Thus, more sustained engagement with local communities is needed to regain local people’s trust, 606 particularly following this series of disappointments (see also Davies & White 2012). It is clear that 607 past outside interventions, which were inadequately planned and did not fulfil their promises, 608 served as a reference point for community members. The long shadow cast by such breaks in trust 609 needs to be taken seriously if sustainable development partnerships are to be built. A scientific 610 agenda alone is thus insufficient for effective conservation. Research in other contexts following 611 sustained experiences of community organising around indigenous planning initiatives, for example 612 by Laurie et al. (2002), indicates that such disappointments influence negotiations over funding, 613 collaboration and investment from outside actors at a community level in unpredictable ways no 614 matter how well-grounded in local needs they are. 615 14

616 5 Conclusions 617 Until now, no empirical research has investigated human relations with peatlands in the Peruvian 618 Amazon. Using an exploratory approach with qualitative empirical research methods (semi- 619 structured interviews; participatory mapping; guided site visits) in two different local communities 620 (one small indigenous community and one comparatively large mestizo community), we found that 621 peatlands are valuable to local people because of their natural resources (palm fruit; wood and 622 timber; game species) and their cultural importance (as important areas for the local mythology), 623 even if ‘peatlands’ per se are not a category of concern to them. Peatlands also occupy a culturally 624 ambiguous position due to the dangers associated with them, such as sinking into the waterlogged 625 ground or being attacked by anacondas and evil spirits. Nevertheless, we also found that the 626 biophysical characteristics of tropical peatlands make them sufficiently distinct to potentially allow 627 peatland-targeted environmental management and conservation activities in collaboration with 628 local communities, who may refer to peatland areas with specific local terms (such as aguajal 629 chupadera, or ‘sucking’ palm swamp; see also Authors 2019). 630 At present, management of these areas is extremely limited as it mostly consists of keeping out 631 non-community members, while the mounting degradation of peatlands was recognised especially 632 by members of the mestizo community who participated in our research. They ascribed degradation 633 to overuse, overpopulation, carelessness, as well as lack of outside support and alternative 634 economic opportunities. 635 It also appeared that ecological degradation is strongly linked with a loss of cultural heritage, with 636 potential implications for peatland conservation. The Urarina still practice the cultural tradition of 637 moving the location of their community every few decades, which allows palm trees, as well as 638 animal populations to recover, but then struggle to have these moves legally recognised by the 639 Peruvian state. Mestizos are instead hoping to modernise their community and switch to a fully 640 monetary and capitalist economy, which seems comparatively less compatible with peatland 641 conservation (even if carbon conservation projects may have potential elsewhere in the Peruvian 642 Amazon, see Roucoux et al. 2017). 643 The main conservation strategies advocated by locals in both communities were limiting access and 644 resource use, sowing new (palm) trees, and above all, creating alternative economic opportunities. 645 These might either be related to peatland use, such as the construction of aguaje fruit processing 646 plants, or unrelated, such as working for oil companies, and in this way, mirror classic debates 647 about the need to combine environmental conservation and economic development. 648 Nominally, most community members approached potential conservation and development 649 strategies with pragmatism. For example, mestizo community members were supportive of the idea 650 of developing markets for sustainable peatland products (see also Roucoux et al. 2017), especially 651 aguaje palm fruit, which could be harvested by climbing palm trees. At present, they are already 652 heavily involved in the unsustainable palm fruit market. Nevertheless, it is also likely that such ideas 653 would face implementation challenges, notably gaining local people’s trust for using novel 654 harvesting techniques after a history of failed development interventions in the area. 655 656 Declaration of interests 657 The authors declare no conflict of interests. 658

15

659 References 660 Álvarez Alonso, J. (2012): Conservación productiva y gestión comunal de la biodiversidad en la 661 Amazonía peruana, in: Martín Brañas, M. (ed.): Articulando la Amazonía: Una Mirada al Mundo 662 Rural Amazónico, Lima, Peru: Agencia Española de Cooperación Internacional para el Desarrollo, 663 7-38. 664 Aquino, R. (2005): Alimentación de mamíferos de caza en los “aguajales” de la Reserva Nacional 665 de Pacaya-Samiria (Iquitos, Perú), in: Revista Peruana de Biología, vol. 12(3): 417-425. 666 Authors (2019), anonymised for peer review: Peatland and wetland ecosystems in Peruvian Amazonia: 667 indigenous classifications and perspectives, unpublished manuscript currently under review. 668 Berkes, F. (2004): Rethinking community-based conservation, in: Conservation Biology, vol. 18(3): 669 621-630. 670 Bhomia, R.K., van Lent, J., Grandez Rios, J.M., Hergoualc’h, K., Honorio Coronado, E.N., & 671 Murdiyarso, D. (2018): Impacts of Mauritia flexuosa degradation on the carbon stocks of 672 freshwater peatlands in the Pastaza-Marañón river basin of the Peruvian Amazon, in: Mitigation 673 and Adaptation Strategies for Global Change, doi:10.1007/s11027-018-9809-9. Epub ahead of print. 674 Boaden, A. (1981): Falcons and falconers: vision in the novels of George Macdonald, in: 675 Christianity & Literature, vol. 31(1): 7-17. 676 Braadbaart, F., Poole, I., Huisman, H.D.J., & van Os, B. (2012): Fuel, fire and heat: an 677 experimental approach to highlight the potential of studying ash and char remains from 678 archaeological contexts, in: Journal of Archaeological Science, vol. 39(4): 836-847. 679 Byg, A., Martin-Ortega, J., Glenk, K., & Novo, P. (2017): Conservation in the face of ambivalent 680 public perceptions – The case of peatlands as ‘the good, the bad and the ugly’, in: Biological 681 Conservation, vol. 206: 181-189. 682 Cheng, A.S. & Mattor, K.M. (2006): Why won’t they come? Stakeholder perspectives on 683 collaborative national forest planning by participation level, in: Environmental Management, vol. 684 38(4): 545-561. 685 Chibnik, M. (1991): Quasi-ethnic groups in Amazonia, in: , vol. 30(2): 167-182. 686 Cossío, R., Menton, M., Cronkleton, P., & Larson, A. (2014): Community Forest Management in the 687 Peruvian Amazon: A Literature Review, Working paper 136, Bogor, Indonesia: CIFOR. 688 Cruickshank, M.M., Tomlinson, R.W., Bond, D., Devine, P.M., & Edwards, C.J.W. (1995): Peat 689 extraction, conservation and the rural economy in Northern Ireland, in: Applied Geography, vol. 690 15(4): 365-383. 691 Dargie, G.C., Lewis, S.L., Lawson, I.T., Mitchard, E.T.A., Page, S.E., Bocko, Y.E., & Ifo, S.A. 692 (2017): Age, extent and carbon storage of the central Congo Basin peatland complex, in: Nature, 693 vol. 542(7639): 86-90. 694 Davies, A.L. & White, R.M. (2012): Collaboration in natural resource governance: reconciling 695 stakeholder expectations in deer management in Scotland, in: Journal of Environmental Management, 696 vol. 112: 160-169. 697 Dean, B. (1994): The poetics of creation: Urarina cosmogony and historical consciousness, in: 698 American Indian Literatures Journal, vol. 10(1): 22-45. 16

699 Dohong, A., Abdul Aziz, A., & Dargusch, P. (2017): A review of the drivers of tropical peatland 700 degradation in South-East Asia, in: Land Use Policy, vol. 69: 349-360. 701 Draper, F.C., Roucoux, K.H., Lawson, I.T., Mitchard, E.T.A., Honorio Coronado, E.N., 702 Lähteenoja, O., Torres Montenegro, L., Valderrama Sandoval, E., Zárate, R., & Baker, T.R. 703 (2014): The distribution and amount of carbon in the largest peatland complex in Amazonia, in: 704 Environmental Research Letters, vol. 9(12): 124017. 705 Endress, B.A., Horn, C.M., & Gilmore, M.P. (2013): Mauritia flexuosa palm swamps: composition, 706 structure and implications for conservation and management, in: Forest Ecology and Management, 707 vol. 302: 346-353. 708 Fabricius, C., Folke, C., Cundill, G., & Schultz, L. (2007): Powerless spectators, coping actors, 709 and adaptive co-managers: a synthesis of the role of communities in ecosystem management, in: 710 Ecology and Society, vol. 12(1): 29. 711 Freitas Alvarado, L., Otárola Acevedo, E., del Castillo Torres, D., Linares Bensimón, C., 712 Martínez Dávila, P., & Malca Salas, G.A. (2006): Servicios Ambientales de Almacenamiento y Secuestro de 713 Carbono del Ecosistema Aguajal en la Reserva Nacional Pacaya Samiria, Loreto - Perú, Documento 714 Técnico no. 29, Iquitos, Peru: Instituto de Investigaciones de la Amazonía Peruana (IIAP). 715 Gapsalamov, A.R. (2015): Fuel industry formation in the region: unknown chapters of history of 716 the Republic of Tatarstan (Russia), in: Research Journal of Applied Sciences, vol. 10(10): 674-679. 717 Garibaldi, A. & Turner, N. (2004): Cultural keystone species: implications for ecological 718 conservation and restoration, in: Ecology and Society, vol. 9(3): 1. 719 Gilmore, M.P. & Young, J.C. (2012): The use of participatory mapping in ethnobiological 720 research, biocultural conservation, and community empowerment: a case study from the Peruvian 721 Amazon, in: Journal of Ethnobiology, vol. 32(1): 6-29. 722 Gonzáles Coral, A. & Torres Reyna, G.M. (2010): Cultivo de Aguaje Mauritia flexuosa L. f.: Manual, 723 Iquitos, Peru: Instituto de Investigaciones de la Amazonía Peruana (IIAP). 724 Halme, K.J. & Bodmer, R.E. (2007): Correspondence between scientific and traditional ecological 725 knowledge: rain forest classification by the non-indigenous ribereños in Peruvian Amazonia, in: 726 Biodiversity and Conservation, vol. 16(6): 1785-1801. 727 Horn, C.M., Vargas Paredes, V.H., Gilmore, M.P., & Endress, B.A. (2018): Spatio-temporal 728 patterns of Mauritia flexuosa fruit extraction in the Peruvian Amazon: implications for 729 conservation and sustainability, in: Applied Geography, vol. 97: 98-108. 730 Ioris, A.A.R. (2017): Places of agribusiness: displacement, replacement, and misplacement in 731 Mato Grosso, Brazil, in: Geographical Review, vol. 107(3): 452-475. 732 Kilbane Gockel, C., & Gray, L.C. (2009): Integrating conservation and development in the 733 Peruvian Amazon, in: Ecology and Society, vol. 14(2): 11. 734 Kramer, B.J. (1979): Urarina Economy and Society: Tradition and Change, PhD thesis, Columbia 735 University, New York, USA. 736 Lähteenoja, O., Rojas Reátegui, Y., Räsänen, M., Del Castillo Torres, D., Oinonen, M., & Page, S. 737 (2012): The large Amazonian peatland carbon sink in the subsiding Pastaza-Marañón foreland 738 basin, Peru, in: Global Change Biology, vol. 18(1): 164-178.

17

739 Laumonier, Y. (1997): The Vegetation and Physiography of Sumatra, Dordrecht, the Netherlands: 740 Kluwer Academic Publishers. 741 Laurie, N., Andolina, R., & Radcliffe, S. (2002): The excluded ‘indigenous’? The implications of 742 multi-ethnic policies for water reform in Bolivia, in: Sieder, R. (ed.): Multiculturalism in Latin 743 America: Indigenous Rights, Diversity and Democracy, Basingstoke, UK: Palgrave Macmillan, 252-276. 744 Lehtinen, A.A. (2000): Mires as mirrors. Peatlands - hybrid landscapes of the North, in: Fennia, 745 vol. 178(1): 125-137. 746 Lilleskov, E., McCullough, K., Hergoualc’h, K., del Castillo Torres, D., Chimner, R., Murdiyarso, 747 D., Kolka, R., Bourgeau-Chavez, L., Hribljan, J., del Aguila Pasquel, J., & Wayson, C. (2018): Is 748 Indonesian peatland loss a cautionary tale for Peru? A two-country comparison of the magnitude 749 and causes of tropical peatland degradation, in: Mitigation and Adaptation Strategies for Global Change, 750 doi:10.1007/s11027-018-9790-3. Epub ahead of print. 751 Medrilzam, M., Smith, C., Abdul Aziz, A., Herbohn, J., & Dargusch, P. (2017): Smallholder 752 farmers and the dynamics of degradation of peatland ecosystems in Central Kalimantan, 753 Indonesia, in: Ecological Economics, vol. 136: 101-113. 754 Messerschmidt, M. (2007): Government and community relations and efforts for comanagement 755 in Macizo de la Muerte, Costa Rica, in: NAPA Bulletin, vol. 27(1): 160-175. 756 Mezzenzana, F. (2018): Encountering Supai: an ecology of spiritual perception in the Ecuadorian 757 Amazon, in: Ethos, vol. 46(2): 275-295. 758 Nath, T.K., Bin Dahalan, M.P., Parish, F., & Rengasamy, N. (2017): Local peoples’ appreciation 759 on and contribution to conservation of peatland swamp forests: experience from Peninsular 760 Malaysia, in: Wetlands, vol. 37(6): 1067-1077. 761 Page, S.E. & Baird, A.J. (2016): Peatlands and global change: response and resilience, in: Annual 762 Review of Environment and Resources, vol. 41: 35-57. 763 Page, S.E., Rieley, J.O., & Banks, C.J. (2011): Global and regional importance of the tropical 764 peatland carbon pool, in: Global Change Biology, vol. 17(2): 798-818. 765 Paniagua-Zambrana, N., Bussmann, R.W., & Macía, M.J. (2017): The socioeconomic context of 766 the use of Euterpe precatoria Mart. and E. oleracea Mart. in Bolivia and Peru, in: Journal of Ethnobiology 767 and Ethnomedicine, vol. 13: 32. 768 Paniagua Zambrana, N.Y., Byg, A., Svenning, J.-C., Moraes, M., Grandez, C., & Balslev, H. 769 (2007): Diversity of palm uses in the western Amazon, in: Biodiversity and Conservation, vol. 16(10): 770 2771-2787. 771 Perú21 (2012): Loreto: Enfrentamiento entre Comuneros y Narcotraficantes Deja un Muerto, available 772 online: https://peru21.pe/lima/loreto-enfrentamiento-comuneros-narcotraficantes-deja-muerto- 773 41546 (last accessed 5 November 2018). 774 Pröpper, M. & Haupts, F. (2014): The culturality of ecosystem services. Emphasizing process and 775 transformation, in: Ecological Economics, vol. 108: 28-35. 776 Ramirez-Gomez, S.O.I., Brown, G., & Tjon Sie Fat, A. (2013): Participatory mapping with 777 indigenous communities for conservation: challenges and lessons from Suriname, in: The Electronic 778 Journal of Information Systems in Developing Countries, vol. 58(1): 2.

18

779 Ricopa Yaicate, R. (2009): Visiones Kukama-Kukamiria en Relación al Bosque y la Sociedad, Serie: 780 Visiones y Conocimientos Indígenas, Iquitos, Peru: AIDESEP, FORMABIAP, ISPPL. 781 Rieley, J. & Page, S. (2016): Tropical peatland of the world, in: Osaki, M. & Tsuji, N. (eds.): 782 Tropical Peatland Ecosystems, Tokyo, Japan: Springer Japan, 3-32. 783 Roucoux, K.H., Lawson, I.T., Baker, T.R., Del Castillo Torres, D., Draper, F.C., Lähteenoja, O., 784 Gilmore, M.P., Honorio Coronado, E.N., Kelly, T.J., Mitchard, E.T.A., & Vriesendorp, C.F. 785 (2017): Threats to intact tropical peatlands and opportunities for their conservation, in: 786 Conservation Biology, vol. 31(6): 1283-1292. 787 Schleicher, J., Peres, C.A., Amano, T., Llactayo, W., & Leader-Williams, N. (2017): Conservation 788 performance of different conservation governance regimes in the Peruvian Amazon, in: Scientific 789 Reports, vol. 7: 11318. 790 Smith, N. (2015): Palms and People in the Amazon, Cham, Switzerland: Springer International 791 Publishing. 792 Staddon, S. (2014): ‘So what kind of student are you?’ The ethics of ‘giving back’ to research 793 participants, in: Lunn, J. (ed.): Fieldwork in the Global South: Ethical Challenges and Dilemmas, London, 794 UK: Routledge, 249-261. 795 Tachibana, T. (2016): Livelihood strategies of transmigrant farmers in peatland of Central 796 Kalimantan, in: Osaki, M. & Tsuji, N. (eds.): Tropical Peatland Ecosystems, Tokyo, Japan: Springer 797 Japan, 613-638. 798 Thorburn, C.C. & Kull, C.A. (2015): Peatlands and plantations in Sumatra, Indonesia: complex 799 realities for resource governance, rural development and climate change mitigation, in: Asia Pacific 800 Viewpoint, vol. 56(1): 153-168. 801 Townsend, J. (1995): Es pot parlar en nom dels altres? Es pot, des de fora, representar les dones 802 pioneres de la selva tropical mexicana? in: Documents d’anàlisi geográfica, vol. 26: 209-218. 803 Walker, H. (2013): Under a Watchful Eye: Self, Power, and Intimacy in Amazonia, Berkeley & Los 804 Angeles, USA: University of California Press. 805 Waylen, K.A., Fischer, A., McGowan, P.J.K., & Milner-Gulland, E.J. (2013): Deconstructing 806 community for conservation: why simple assumptions are not sufficient, in: Human Ecology, vol. 807 41(4): 575-585. 808 Wilson, A. (2018): Swamp: Nature and Culture, London, UK: Reaktion Books. 809 Young, J.C. & Gilmore, M.P. (2013): The spatial politics of affect and emotion in participatory 810 GIS, in: Annals of the Association of American Geographers, vol. 103(4): 808-823. 811 Zinngrebe, Y. (2016): Incorporating Biodiversity Conservation in Peruvian Development – A History with 812 Different Episodes, Discussion paper no. 1606, Department for Agricultural Economics and Rural 813 Development, Georg-August-Universität Göttingen, Germany.

19

814 815 Figure 1: (a) Location of the Pastaza-Marañón Basin in western Amazonia (dashed box); (b) inset showing the Pastaza- 816 Marañón Basin, with the modelled distribution of peatlands following Draper et al. (2014) in grey. The study area, around 817 the Chambira and Tigre River Basins of Peru’s Loreto Region, is indicated. AdP stands for ‘Abanico del Pastaza’, a Ramsar 818 wetland site; PSNR stands for Pacaya-Samiria National Reserve, a protected area.

819 Table 1: Summary of uses, cultural significance, and current management of peatlands in two communities of the Peruvian 820 Amazon (explained in depth in sections 3.1-3.3) Mestizo community Indigenous community (Urarina) Local terms likely Aguajal chupadera (‘sucking’ Alaka (palm swamp forest; associated with peatland palm swamp dominated by permanently to seasonally wet areas Mauritia flexuosa) ecosystem) Aguajal raizal/champal (palm swamp with comparatively firm ground dominated by Mauritia flexuosa) Aguajal varillal (Mauritia flexuosa Jiiri (‘open space’; permanently palm swamp with short and wet ecosystem including open thin trees like pole forest) peatland areas and pole forest) Piripiral (sedgeland) “Allá es feísimo.” (“very ugly” areas) Uses (section 3.1) Palm fruit for own Palm fruit for own consumption (Mauritia flexuosa; consumption (Mauritia flexuosa; Oenocarpus batahua; Mauritiella Oenocarpus batahua; Mauritiella armata) armata)

Palm fruit for trade (Mauritia Palm fruit for small-scale trade flexuosa), including intra- (Mauritia flexuosa) with community intermediaries and travelling traders travelling traders

20

Palm leaves (Attalea butyracea) Palm leaves (Attalea butyracea) for roofing for roofing

Palm hearts for trade (Euterpe Palm hearts for small-scale precatoria), including intra- trade (Euterpe precatoria) with community intermediaries and travelling traders travelling traders Palm fibre for textile production (Mauritia flexuosa)

Wood and timber for personal Wood and timber for personal use and trade (e.g. Virola sp.; use (mostly Virola sp.) Lauraceae; Hevea brasiliensis, Calophyllum brasiliense) Hunting (e.g. tapirs, caimans, Hunting (e.g. tapirs, caimans, peccaries, agoutis, various peccaries, agoutis, various monkey species, Spix’s guan, monkey species, Spix’s guan, great tinamou) great tinamou) Cultural significance Home of various locally Jiiri and alaka as home of the (section 3.2) confined guardian spirits (but Baainu, a dangerous guardian not exclusive to peatlands) spirit Area where ‘dead lakes’ can be Area where Mauritia flexuosa found (cultural taboo areas grows, with importance for where natural resource use Urarina creation myth (Dean carries risks for one’s life) 1994) Current management and Local community self- Local community self- conservation (section 3.3) governance with dual system of governance with dual system of indigenous and state- indigenous and state- recognised authorities, and recognised authorities, and community assemblies community assemblies Safeguarding from resource use Safeguarding from resource use by non-community members by non-community members Trade limits for fish; fixed Ban on timber harvesting and prices for intra-community use of fish poison, mostly trade of fish and meat; very complied with limited compliance Strong sense of resource Limited sense of resource depletion, including game depletion, including game species, palm trees, timber species, palm trees, timber Individual adoption of Limited adoption of sustainable sustainable practices, such as practices, such as planting of climbing Mauritia flexuosa palm tree seedlings (with NGO trees, planting Mauritia flexuosa support) seedlings, alternative monetary income strategies Community has been in the Community has been in the same place since foundation in same place since the 1980s, 1933, area of depleted change of community location resources is gradually every few decades to allow expanding outwards resource regeneration 821 21